MFBO-SSM: Multi-Fidelity Bayesian Optimization for Fast Inference in State-Space Models

Authors: Mahdi Imani, Seyede Fatemeh Ghoreishi, Douglas Allaire, Ulisses M. Braga-Neto7858-7865

AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The accuracy and speed of the algorithm are demonstrated by numerical experiments using synthetic gene expression data from a gene regulatory network model and real data from the VIX stock price index.
Researcher Affiliation Academia Mahdi Imani Texas A&M University m.imani88@tamu.edu Seyede Fatemeh Ghoreishi Texas A&M University f.ghoreishi88@tamu.edu Douglas Allaire Texas A&M University dallaire@tamu.edu Ulisses M. Braga-Neto Texas A&M University ulisses@ece.tamu.edu
Pseudocode Yes Algorithm 1 MFBO-SSM Algorithm
Open Source Code No The paper does not provide an explicit statement or link to open-source code for the methodology.
Open Datasets No The paper mentions using
Dataset Splits No The paper describes data generation and lengths (e.g.,
Hardware Specification Yes All experiments have been conducted on a PC with an Intel Core i7-4790 CPU@3.60-GHz clock and 16 GB of RAM.
Software Dependencies No The paper does not specify software names with version numbers for its dependencies.
Experiment Setup Yes MFBO-SSM algorithm uses N1 = 100, N2 = 1000, N3 = 5000, corresponding to small, medium, and large particle sample sizes. Other methods use a fixed particle sample size N = 1000. [...] We are interested in estimating the true parameter θ = (σ , φ , β , µ ) = (0.97, 0.55, 0.95, 0.1) from synthetic data, where Θ = [0, 2] [ 1, 1] [0, 10] [0, 5]. [...] The MFBO-SSM, BO, EM, and ML algorithms all stop when the change in the estimated value of all parameters over a window of length 20 falls bellow 5% of their range, whereas the PMMH algorithm continues over a fixed number of 6,000 iterations.